When a big disaster occurs, a rescue team has to send the food and/or supplies to an unspecified number of affected people. Therefore, an evolutionary method that can provide service to unspecific large number of users is needed. In this paper, an evolutionary method, Self-Administered Genetic Algorithm (SAGA) is proposed. This method applies systems of intelligent autonomous agents that are operated on a block chain network and evaluated by cryptocurrency. The intelligent autonomous agents can operate on themselves by spending their cryptocurrency to a node of the block chain. The intelligent autonomous agents (parents) increase by creating copies (children) of themselves. The children initially acquire cryptocurrency from their parents. As a result, the intelligent autonomous agents can evolve autonomously in SA-GA. SA-GA can be a method that optimizes a system consisting of a large number of autonomous agents1.